Method and System for Next Word Prediction

1. A method implemented on a computing device having at least one processor, storage, and a communication platform connected to a network for predicting a next word, the method comprising: generating a suggestion database including a plurality of suggestions; receiving an input from a user including a first word and at least one character of a second word; obtaining one or more indices associated with the at least one character from the suggestion database; retrieving one or more suggestions for the second word from each of the one or more indices based on a frequency with which each of the one or more suggestions for the second word were previously retrieved for one or more other users to determine a ranking of the retrieved one or more suggestions for the second word; determining at least one suggestion from the one or more suggestions based on a measure with respect to the first word and each of the one or more suggestions and the ranking; and providing the at least one suggestion to the user for presentation.

2. The method of claim 1 , wherein generating the plurality of suggestions further comprises: obtaining training data related to content suggestions; generating a plurality of strings based on the training data, each of the plurality of strings comprising at least two words of suggestion; expanding a last word of each of the plurality of strings to obtain a set of items, wherein each item of the set of item is a partial prefix of the last word; and predicting one or more next words based on the set of items, each of the one or more next words being a suggestion of the last word.

3. The method of claim 2 , wherein predicting the one or more next words based on the set of items further comprises: predicting one or more candidate words that match the partial prefixes of the last word; ranking the one or more candidate words based on at least a first criteria; and filtering the one or more ranked candidate words based on a predetermined threshold to generate the one or more next words.

4. The method of claim 2 , wherein a dimension of the set of items is equal to a number of characters of the last word, and each item of the set of items has one character more than a previous item in the set of items.

5. The method of claim 2 , further comprising: generating an index for each of the one or more next words; generating a prefix tree for each of the plurality of strings comprising partial prefixes of the last word; and storing the index for each of the one or more next words and the prefix trees in the suggestion database.

6. The method of claim 5 , further comprising: associating additional information with each of the one or more next words; and storing the additional information with the one or more next words in the suggestion database.

7. The method of claim 1 , wherein the ranking of the retrieved one or more suggestions for the second word is determined based on the measure, and wherein determining the at least one suggestion from the one or more suggestions comprises: selecting the at least one suggestion from the ranked one or more suggestions based on the measure, wherein the measure comprises a frequency with which a second word corresponding to a given suggestion of the one or more suggestions is recommended with the first word.

8. A system having at least one processor, storage, and a communication platform for predicting a next word comprising: a suggestion database generating engine implemented on the at least one processor and configured to generate a suggestion database including a plurality of suggestions; a user interface implemented on the at least one processor and configured to receive an input from a user including a first word and at least one character of a second word; an index retriever implemented on the at least one processor and configured to obtain one or more indices associated with the at least one character from the suggestion database; a suggestion retriever implemented on the at least one processor and configured to: retrieve one or more suggestions for the second word from each of the one or more indices from the suggestion database based on a frequency with which each of the one or more suggestions for the second word were previously retrieved for one or more other users to determine a ranking of the retrieved one or more suggestions for the second word, and determine at least one suggestion from the one or more suggestions based on a measure with respect to the first word and each of the one or more suggestions and the ranking; and a presenting module implemented on the at least one processor and configured to provide the at least one suggestion to the user for presentation.

9. The system of claim 8 , wherein the suggestion database generating engine further comprises: a string builder configured to: obtain training data related to content suggestions, and generate a plurality of strings based on the training data, each of the plurality of strings comprising at least two words of suggestion; a word expanding module configured to expand a last word of each of the plurality of strings to obtain a set of items, wherein each item of the set of item is a partial prefix of the last word; and a next word predicting module configured to predict one or more next words based on the set of items, each of the one or more next words being a suggestion of the last word.

10. The system of claim 9 , wherein: the next word predicting module is further configured to predict one or more candidate words that match the partial prefixes of the last word, and the suggestion database generating engine further comprises: a ranking module configured to rank the one or more candidate words based on at least a first criteria; and a filtering module configured to filter the one or more ranked candidate words based on a predetermined threshold to generate the one or more next words.

11. The system of claim 9 , wherein a dimension of the set of items is equal to a number of characters of the last word, and each item of the set of items has one character more than a previous item in the set of items.

12. The system of claim 9 , further comprising: a suggestion generating module configured to generate the plurality of suggestions based on the one or more next words, wherein the suggestion generating module further comprises: an indexing unit configured to generate an index for each of the one or more next words; a prefix tree generating unit configured to generate a prefix tree for each of the plurality of strings comprising partial prefixes of the last word; and a storing unit configured to store the index for each of the one or more next words and the prefix trees in the suggestion database.

13. The system of claim 12 , wherein the suggestion generating module further comprises: a suggestion associating unit configured to associate additional information with each of the one or more next words; and the storing unit is further configured to store the additional information with the one or more next words in the suggestion database.

14. The system of claim 8 , further comprising: a ranking module configured to determine the ranking of the retrieved one or more suggestions based on the measure, wherein the at least one suggestion being determined from the one or more suggestions comprises the at least one suggestion being selected from the ranked one or more suggestion based on the measure, wherein the measure comprises a frequency with which a second word corresponding to a given suggestion of the one or more suggestions is recommended with the first word.

15. A non-transitory machine-readable medium having information recorded thereon for predicting a next word, when read by the machine, causes the machine to perform operations comprising: generating a suggestion database including a plurality of suggestions; receiving an input from a user including a first word and at least one character of a second word; obtaining one or more indices associated with the at least one character from the suggestion database; retrieving one or more suggestions for the second word from each of the one or more indices from the suggestion database based on a frequency with which each of the one or more suggestions for the second word were previously retrieved for one or more other users to determine a ranking of the retrieved one or more suggestions for the second word; determining at least one suggestion from the one or more suggestions based on a measure with respect to the first word and each of the one or more suggestions and the ranking; and providing the at least one suggestion to the user for presentation.

16. The medium of claim 15 , wherein the operations further comprise: obtaining training data related to content suggestions; generating a plurality of strings based on the training data, each of the plurality of strings comprising at least two words of suggestion; expanding a last word of each of the plurality of strings to obtain a set of items, wherein each item of the set of item is a partial prefix of the last word; and predicting one or more next words based on the set of items, each of the one or more next words being a suggestion of the last word.

17. The medium of claim 16 , wherein the operations further comprise: predicting one or more candidate words that match the partial prefixes of the last word; ranking the one or more candidate words based on at least a first criteria; and filtering the one or more ranked candidate words based on a predetermined threshold to generate the one or more next words.

18. The medium of claim 16 , wherein a dimension of the set of items is equal to a number of characters of the last word, and each item of the set of items has one character more than a previous item in the set of items.

19. The medium of claim 16 , wherein the operations further comprise: generating an index for each of the one or more next words; generating a prefix tree for each of the plurality of strings comprising partial prefixes of the last word; and storing the index for each of the one or more next words and the prefix trees in the suggestion database.

20. The medium of claim 16 , wherein the operations further comprise: associating additional information with each of the one or more next words; and storing the additional information with the one or more next words in the suggestion database.